One important characteristic of the normal human cognitive system is its ability to monitor its own behavior. When a particular task must be accomplished, links among components of the cognitive system are established and the dynamics of the system are adjusted so as to satisfy the goals of the task. The performance of the system is then evaluated within the framework of these goals. In this context, monitoring involves a comparison between actual performance and some desired performance, with performance being defined in terms of the speed and/or accuracy of a response. A related aspect of the cognitive system is its ability to initiate remedial action when the monitoring system detects a discrepancy between actual and desired performance (i.e., an error). Possible actions include immediate attempts to inhibit or correct the error, or longer term (re)-adjustments that will reduce the likelihood of future errors. The proposed research will focus on this ability of the human cognitive system to monitor its own behavior and to initiate remedial action of the behavior does not conform to task goals. The research will utilize the psychophysiological approach in which the traditional measures of mental chronometry (speed and accuracy) are augmented by measures of psychophysiological function (e.g., scalp-recorded event-related brain potentials and the electaromyogram). Particular emphasis will be given to measures of a component of the event-related potential (the error-related negativity or ERN) that is specifically related to error-processing. The ERN can be detected by analyzing electrical brain activity that is time- locked to the execution of an incorrect response. The research will elucidate the error-related processes that are implemented by the neural system that underlies the ERN, and will use this information to elaborate on error-detection and related mechanisms. In particular, the research will evaluate the generality of the error-detection process, the nature of the representations used by the error-detection process, the automaticity of error-related processes, and the degree to which connectionist models of error processing can approximate actual data. The research is important because it deals with an aspect of human cognitive function that is dysfunctional in some psychiatric and neurological disorders, but that has been neglected in most prior research.